{"id":"https://openalex.org/W4411280530","doi":"https://doi.org/10.1145/3699682.3728347","title":"Task-specific, personalized Automatic Speech Recognition","display_name":"Task-specific, personalized Automatic Speech Recognition","publication_year":2025,"publication_date":"2025-06-13","ids":{"openalex":"https://openalex.org/W4411280530","doi":"https://doi.org/10.1145/3699682.3728347"},"language":"en","primary_location":{"id":"doi:10.1145/3699682.3728347","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3699682.3728347","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3699682.3728347","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 33rd ACM Conference on User Modeling, Adaptation and Personalization","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://dl.acm.org/doi/pdf/10.1145/3699682.3728347","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5053335512","display_name":"Fahrettin G\u00f6kg\u00f6z","orcid":"https://orcid.org/0000-0002-2927-3841"},"institutions":[{"id":"https://openalex.org/I4210166245","display_name":"Fraunhofer Institute for Communication, Information Processing and Ergonomics","ror":"https://ror.org/05nn0gw40","country_code":"DE","type":"facility","lineage":["https://openalex.org/I4210166245","https://openalex.org/I4923324"]}],"countries":["DE"],"is_corresponding":true,"raw_author_name":"Fahrettin G\u00f6kg\u00f6z","raw_affiliation_strings":["Fraunhofer FKIE, Wachtberg, Germany"],"raw_orcid":"https://orcid.org/0000-0002-2927-3841","affiliations":[{"raw_affiliation_string":"Fraunhofer FKIE, Wachtberg, Germany","institution_ids":["https://openalex.org/I4210166245"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5051815123","display_name":"Hussein Hasso","orcid":null},"institutions":[{"id":"https://openalex.org/I4210166245","display_name":"Fraunhofer Institute for Communication, Information Processing and Ergonomics","ror":"https://ror.org/05nn0gw40","country_code":"DE","type":"facility","lineage":["https://openalex.org/I4210166245","https://openalex.org/I4923324"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Hussein Hasso","raw_affiliation_strings":["Fraunhofer FKIE, Wachtberg, Germany"],"raw_orcid":"https://orcid.org/0009-0001-1673-9964","affiliations":[{"raw_affiliation_string":"Fraunhofer FKIE, Wachtberg, Germany","institution_ids":["https://openalex.org/I4210166245"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":["https://openalex.org/A5053335512"],"corresponding_institution_ids":["https://openalex.org/I4210166245"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.06283594,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"194","last_page":"203"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":0.9998999834060669,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":0.9998999834060669,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10860","display_name":"Speech and Audio Processing","score":0.9991000294685364,"subfield":{"id":"https://openalex.org/subfields/1711","display_name":"Signal Processing"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11309","display_name":"Music and Audio Processing","score":0.9988999962806702,"subfield":{"id":"https://openalex.org/subfields/1711","display_name":"Signal Processing"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7913967370986938},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.6490129232406616},{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.5979142189025879},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.4581325352191925},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.4014959931373596},{"id":"https://openalex.org/keywords/engineering","display_name":"Engineering","score":0.06230232119560242}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7913967370986938},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.6490129232406616},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.5979142189025879},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.4581325352191925},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4014959931373596},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.06230232119560242},{"id":"https://openalex.org/C201995342","wikidata":"https://www.wikidata.org/wiki/Q682496","display_name":"Systems engineering","level":1,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1145/3699682.3728347","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3699682.3728347","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3699682.3728347","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 33rd ACM Conference on User Modeling, Adaptation and Personalization","raw_type":"proceedings-article"},{"id":"pmh:oai:publica.fraunhofer.de:publica/503105","is_oa":true,"landing_page_url":"https://publica.fraunhofer.de/handle/publica/503105","pdf_url":null,"source":{"id":"https://openalex.org/S4306400318","display_name":"Fraunhofer-Publica (Fraunhofer-Gesellschaft)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I4923324","host_organization_name":"Fraunhofer-Gesellschaft","host_organization_lineage":["https://openalex.org/I4923324"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"conference paper"}],"best_oa_location":{"id":"doi:10.1145/3699682.3728347","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3699682.3728347","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3699682.3728347","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 33rd ACM Conference on User Modeling, Adaptation and Personalization","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4411280530.pdf","grobid_xml":"https://content.openalex.org/works/W4411280530.grobid-xml"},"referenced_works_count":28,"referenced_works":["https://openalex.org/W1538019985","https://openalex.org/W1970491336","https://openalex.org/W2124509324","https://openalex.org/W2127141656","https://openalex.org/W2127851351","https://openalex.org/W2407080277","https://openalex.org/W2591804103","https://openalex.org/W2612690371","https://openalex.org/W2936774411","https://openalex.org/W2963057973","https://openalex.org/W2990138404","https://openalex.org/W3010951478","https://openalex.org/W3015216127","https://openalex.org/W3016008406","https://openalex.org/W3096947210","https://openalex.org/W3112702554","https://openalex.org/W3161480375","https://openalex.org/W3196475561","https://openalex.org/W3209984917","https://openalex.org/W3215615641","https://openalex.org/W4200300291","https://openalex.org/W4296906760","https://openalex.org/W4385823256","https://openalex.org/W4390075359","https://openalex.org/W4392745070","https://openalex.org/W4399372393","https://openalex.org/W4401610398","https://openalex.org/W4406107190"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W4391913857","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W3204019825"],"abstract_inverted_index":{"Voice":[0],"User":[1],"Interfaces":[2],"(VUI)":[3],"are":[4,113],"particularly":[5],"useful":[6],"if":[7,15,111],"the":[8,23,27,34,36,49,54,61,159,162,165,213,216,225,232],"operator":[9,28],"has":[10],"to":[11,53,115,129,139,164,185,199,206,230],"work":[12],"hands-free":[13],"or":[14,45],"her":[16],"cognitive":[17],"load":[18],"is":[19,22,43,204,228],"very":[20],"high.This":[21],"case,":[24],"e.g.,":[25],"when":[26],"can":[29,107],"be":[30,67,108],"easily":[31],"disturbed":[32],"by":[33],"environment,":[35],"operational":[37,167],"task":[38],"induces":[39],"stress":[40],"and":[41,76,87,120,150,161,237],"there":[42],"little":[44],"no":[46],"fault":[47],"tolerance.However,":[48],"factors":[50],"that":[51,178],"contribute":[52],"usefulness":[55],"of":[56,80,219,234],"a":[57,127,131,146,195],"VUI":[58],"also":[59],"complicate":[60],"design.Automatic":[62],"Speech":[63,135,200],"Recognition":[64],"(ASR)":[65],"must":[66],"robust":[68,110],"in":[69],"noisy":[70],"environments,":[71],"under":[72,102],"non-optimal":[73],"microphone":[74],"conditions":[75],"for":[77,122],"different":[78],"types":[79],"speech":[81,210],"-including":[82],"stress-induced":[83],"shouting,":[84],"hyper":[85],"articulation":[86],"heavy":[88],"breathing,":[89],"among":[90],"others.Commercially":[91],"available,":[92],"generic":[93],"ASR":[94,105],"solutions":[95],"do":[96],"not":[97],"fulfil":[98],"high":[99],"robustness":[100],"requirements":[101],"these":[103,222],"conditions.However,":[104],"systems":[106],"made":[109],"they":[112],"tailored":[114],"their":[116,190],"respective":[117],"use":[118],"cases":[119],"personalised":[121],"specific":[123,166],"users.This":[124],"paper":[125],"introduces":[126],"method":[128],"customize":[130],"Large":[132],"Vocabulary":[133],"Continuous":[134],"Recognizer":[136],"(LVCSR)":[137],"system":[138,144,227],"achieve":[140],"such":[141],"robustness.An":[142],"LVCSR":[143,226],"includes":[145],"language":[147],"model":[148,153,214],"(LM)":[149],"an":[151,173,180],"acoustic":[152],"(AM).The":[154],"customization":[155],"involves":[156],"adapting":[157],"both":[158],"LM":[160,169],"AM":[163,193],"context.For":[168],"customization,":[170,194],"we":[171],"employ":[172],"Use":[174],"Case":[175],"Editor":[176],"(UCE)":[177],"provides":[179],"intuitive":[181],"interface,":[182],"enabling":[183],"users":[184],"align":[186],"linguistic":[187],"models":[188],"with":[189],"unique":[191],"needs.For":[192],"Multi":[196],"Speaker":[197],"Text":[198],"Synthesis":[201],"(MSTTS)":[202],"module":[203],"used":[205],"automatically":[207],"generate":[208],"personalized":[209],"data,":[211],"ensuring":[212],"captures":[215],"distinctive":[217],"characteristics":[218],"individual":[220],"speakers.Together,":[221],"adaptations":[223],"ensure":[224],"configured":[229],"meet":[231],"demands":[233],"challenging":[235],"environments":[236],"diverse":[238],"users.":[239]},"counts_by_year":[],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}